The overall goal of this proposal is to determine if molecular markers found in pretreatment biopsy specimens can predict pathologic complete response (pathCR) in patients with esophageal cancer (EC) undergoing preoperative chemoradiation (CTRT). PathCR is defined as no residual carcinoma in a resected specimen. Patients with EC have a 5-year survival rate of <20%. Localized carcinoma (stage II or III) is frequently treated with preoperative CTRT, but this empirical approach results in heterogeneous and often unpredictable outcomes. Approximately 25% of patients achieve a pathCR and tend to survive longer than do the 75% who achieve <pathCR (i.e., have residual cancer). Current clinical parameters do not reliably predict outcome from therapy and modifying the type CTRT has little impact on the proportion of patients with a pathCR. Moreover, preoperative CTRT and surgery have dire toxicity and life altering consequences. This lack of ability to individualize therapy is a major obstacle in improving patient outcomes. A rational approach to individualized therapy is likely to occur through understanding the molecular biology of this cancer. To that aim, we conducted a preliminary study in EC patients. In a 19-patient pilot profiling study using Affymetrix U133A GeneChip. microarray, unsupervised hierarchical cluster analysis segregated patient's cancers into two subgroups (5 of 6 pathCR clustered in subtype I and 1 of 6 in subtype II). These data suggest that pathCR might segregate from <pathCR based on molecular markers but our data needs confirmation in a larger patient sample. In an expanded cohort, we may also find an improved model for predicting pathCR which would itself require validation. We propose to assess the gene expression profile of 120 patients to assess if a high (e80%) positive predictive value for pathCR can be achieved. The Specific Aims of the proposal are 1. Determine the gene expression profile of the new cohort of 120 patients with esophageal cancer treated with preoperative chemoradiation. 2: Construct statistical models that have a high positive predictive value for pathCR by the gene expression profile of 120 patients. The modeling will include: (a) Identification of subtypes of esophageal cancer by gene expression profiles that are highly homogeneous; (b) Build pathCR prediction models based on specific to subtypes of esophageal cancer as defined in 2(a) and also in the homogeneously treated subgroups. Our attempts to individualize patient's therapy based on molecular biology can pave the way to allow administration of effective therapy, improve safety, and preserve the esophagus in some patients. PUBLIC HEALTH RELEVANCE: This proposal is an early attempt to individualize therapy based on molecular biology for patients with esophageal cancer. Our goal is to pave the way for a strategy in the future that will allow administration of effective therapy, improve safety, and preserve the esophagus in some patients. [unreadable] [unreadable] [unreadable]